Full text of "USPTO Patents Application 09843106"

See other formats

Europaisches Patentamt
European Patent Office
Office europeen des brevets

@ Application number : 94307596.0
@ Date of filing : 17.10.94

(So) Priority : 15.10.93 US 137880

@ Date of publication of application :
19.04.95 Bulletin 95/16

(&) Designated Contracting States :

AT BE CH DE ES FR GB IT U NL SE

(7j) Applicant : International Business Machines
Corporation
Old Orchard Road
Armonk, N.Y. 10504 (US)

@ Inventor: Tsevdos, James T.
2711 NE 57th Street
Fort Lauderdale, Florida 33308 (US)
Inventor : Cook, Ross L.
901 SW 36th Avenue
Boynton Beach, Florida 33435 (US)
Inventor : Ring, Nancy Lee
6129 Town Colony Drive
Boca Raton, Florida 33433 (US)

Inventor : Barnhill, Robert S.
21218 St Andrews Blvd.,
No. 10-405

Boca Raton, Florida 33433 (US)

Inventor : Hamblin, Glen E.

9384 Gettysburg Road

Boca Raton, Florida 33434 (US)

Inventor : Mil stead, Kenneth l_

9927 Majestic Way

Boynton Beach, Florida 33437 (US)

Inventor : Kindell, Craig N.

2014 SW 29th Court No. 2A

Delray Beach, Florida 33445 (US)

Inventor : Waefler, Susan Elizabeth

5086 Madison Road

Delray Beach, Florida, 33484 (US)

Inventor : Portela, Carlos

2659 SE 14th Street

Pompano Beach, Florida 33062 (US)

Inventor : Anderson, Brent C.

14232 Marsh Lane

Dallas, Texas 75244 (US)

(g) Representative : Burt, Roger James, Dr.
IBM United Kingdom Limited
Intellectual Property Department
Hursiey Park

Winchester Hampshire S021 2JN (GB)

CM
<

CL
UJ

(54) Digital information accessing, delivery, and reproduction.

@ A digital data on-demand turnkey system at a customer premise wherein N number of servers provide
for 100% of content distribution of remotely stored digitized information, which information may be
previewed in real-time, and product incorporating selected digitized information can be manufactured
on-site and within a short response time to a customer's request at a point of sale location. In a retail
environment customers, at a point-of-sale location are able to exhaustively search and preview the
content database using graphics-based touch screens at consumer kiosks. Previews including audio
and video segments are made available. Prompting screens allow customers to make purchasing
decisions by stipulating content which is available from any number of categories of subject matter
including music. Content can be identified by album, artist or partial portions of material within the
content of the totality of the work. The selected media for the manufacture and production of the digital
data may be from a myriad of different selections and can include CD's, cassette tapes, CD ROM
technology, reel-to-reel tapes, and video disks, as an example. Since the system is adaptable to having
an unlimited number of servers, strategic and economic networking is accomplished to allocate
proportionally the preview material as well as the total content of a specific work at the respective
servers depending, in particular, on the popularity of the content and the costs of transmission. A
master server will be situated geographically so as to be accessible to chain and network subservers. Its
geographic and networked location is dependent upon communication network systems and subsys-
tem costs and availability in order to best serve a customer's premise, whether it be a retail store or
similar point-of-sale or other end-user location. The strategy of allocating the content between the
servers, preview and for the final manufacturing process for a particular work, is done so as to eliminate
the need for any in-store inventory of, for example, an artist's composition regardless of format. Security
mechanisms that require centralized database authorizations prior to the transmission of content
and/or the manufacture of any of the products is provided in addition. The systems provides simplified
financial and accounting services. The total authorization, accounting and management reporting and

Jouve, 18, rue Saint-Denis, 75001 PARIS

EP 0 649 121 A2

marketing system collects and maintains data useful to the retailer and the content data owners or
managers, such as label companies.

I32

LOCAL SITES ^"6 jlQ2 MASTER SITE

DATA CAPTURE
PROCESS

• DIRECTORY SVCS h
♦OAT A SERVER

• NETWORK MGMT
•SWITCH/HUB/ADP

REGIONAL
SITES

• DIRECTORY SVCS h

• DATA SERVER

• NETWORK MGMT

• ROYALTY ACCT'G

• SWITCH/HUB/

ADP

is;

NETWORK
MGMT

OIRECTORY
SVCS

DATA

SERVER OPTICAL
JUKEBOXES

O B O O 0

MULTIMEDIA PREVIEW STATIONS

COLOR PRINTER

FIG. I

EP 0 649 121 A2

BACKGROUND OF THE INVENTION
Field of the Invention

5 The present invention relates to the field of data storage and data communication. In particular, the inven-

tion relates to a system for accessing digital information, such as audio and video information, at remotely stor-
ed locations and for communicating in real-time that information to another location (such as a user's premise
or a point-of-sale) for reproduction of the communicated information in a material object for end-user use.

10 Description of Related Art

Retailers of the world are in a constant struggle to meet the demands of the consuming public with regard
to the availability and competitive pricing of products. This struggle is a challenge for the retailer's ability to
meet the end-user's immediate need and if an item must be backordered, the customer's patience. The retailer

15 must do that in a manner which is cost effective to the merchandiser and competitive from a delivery and price
standpoint to the purchaser.

There is likely no more of a demanding and aggressive purchasing public than those individuals who are
avid followers and fans of entertainers and artists in the entertainment industry. Because of the high demand
and the massive number of customers and potential customers, as it relates to the entertainment industry, the

20 level of competition is extremely high, whether it be for memorabilia, contemporary news items, fashion trends,
or entertainment products.

Product distribution and availability is one of those merchandising areas where the retailer has experi-
enced keen competition and where they have struggled to be the source of all products, to be identified as
having immediate availability, to continue to serve the totality of the market share which they have captured

25 and to grow that market share.

As an example, in the entertainment industry, particularly as it applies to companies involved in music typ-
ically known as music label companies, there has been a constant challenge for such companies to exploit their
entire catalog. There is a need or concern to make every album in their catalog available in any format and to
never be out of stock with respect to that particular product. If the labels are indeed able to achieve that high

30 level of expectation and availability, the retailer then is faced with the dilemma of allocating floor space to all
of that available inventory regardless of the popularity of much of the stocked material. If this is not common
practice, the retailer will be faced with losing a particular customer and clientele if the stock is not on-hand. In
most instances, it is unrealistic to expect that a retailer will have sufficient floor space and financial resources
to have an exhaustive stock of all products. As an example, the music industry is presently offering 60,000 to

35 100,000 albums by title and is growing at an extraordinary rate. Additionally, artists are continually engaged
to expand their repertoire. This is particularly true when an artist has gained recent popularity and is a big hit
with the consuming public, which may have resulted in an insatiable appetite for that artist's work. The retailer
has the challenge of identifying what quantities are marketable over what period of time for that artist and any
number of other artists and labels. The outlet merchandiser is fraught with the effort to stock those products

40 and to have them available when approached by a customer who has that immediate and demanding interest
in procuring a copy of that artist's work. Any miscalculation by the retailer evolves in what is known as stock-
outs.

The retailer's inventory management problems are there whether or not the product is in inventory at their
particular retail outlet or needs to be procured in order to meet that immediate demand regardless of the life

45 of the demand. These are time consuming and difficult calculations for retailers. The retailer break-even point
is in a constant state of flux and miscalculations and bad judgment calls result in prices which are less than
competitive and margins which are often nonexistent.

A further frequently encountered problem by the retailer is the issue of a purchaser who may not be in the
position to fully identify either the artist by full name, the label company or the full name of the title of the auth-

50 ored work or piece. Now with respect to items which are of current popularity, some developed systems are
available to preview portions that may be able to aid in the Identification of the particular product of interest.
However, when it comes to the literally hundreds of thousands of pieces that have been produced, it is a near
impossibility under the present systems to assist that potential purchaser when that artistic production is not
of current vintage and even may be extremely dated. Further, even if some assistance has been given to that

55 particular purchaser, without preview of that item, there is a great deal of uncertainty on the part of the potential
purchaser as to whether or not the item identified is one of interest to them. In most cases there is great hes-
itancy to purchase an item which may be relatively expensive and which may turn out to not be a item of par-
ticular interest.

EP 0 649 121 A2

It is difficult for the retailer to maximize the inventory at the retail outlet and have it attain the level of ex-
pectation of the customer. Inventory and product stocking principles, such as just-in-time practices, have as-
sisted the merchandiser to control the amount of stock-on-hand against cash-flow concerns to meet customer
demand. But these practices still fall short of the inpatient consumer who wants the product now regardless
5 of the reasonableness of the demand, even for products which are so popular that stock-outages are to be
expected or products which are so dated that only archival copies generally exist.

Even in today's automated systems which provide a computerized library resource of available program-
med materials, limitations are imposed and financial constraints are present due to a lack of convenient and
available storage space and constricted and time consuming data communication channels.

SUMMARY OF THE INVENTION

In one aspect, the present invention provides a method for accessing digital information stored in a source
library at a first location and recording a portion of the digital information, which information is processed for
is subsequent playback, onto a recordable medium at a location other than the first location, the method including
the steps of: identifying, at a second location remote from the first location, a portion of the digital information
in the source library; sending a request from the second location to the source library for the identified portion
of the digital information in the source library to be transferred from the source library; validating that the iden-
tified portion of the digital information is authorised for playback; transferring in real-time information within
20 the identified portion of the digital information to a manufacturing device at a location other than the first lo-
cation for. incorporation of the digital information on a medium suitable for playback; and incorporating the iden-
tified portion of the digital information on the medium in real-time in a manner suitable for playback.

In a second aspect, the invention provides a method of allocating for storage 100 percentage of content
data amongst a plurality of data servers remotely disposed to each other and connected in hierarchal levels
25 having a low and high end and in data communication with a data accessing station also remotely disposed
from each of the plurality of data servers, but less so disposed from the data server at the low end of the hi-
erarchy than from each succeeding data server at the next succeeding higher level of the hierarchy through
and to the data server at the high end of the hierarchy of data servers, comprising the steps of: allocating for
storage X percentage of the content data at the data server at the low end of the hierarchy least remote from
30 the data accessing station which X percentage of allocated content data is a function of content data for most
frequent accessing by the data accessing station; allocating for storage Y percentage, where Y is < 100 and
> X, of the content data at the data server at the high end of the hierarchy of data servers most remote from
the data accessing station as a function of content data for least frequent accessing by the data accessing
station; and allocating for storage 100-X-Y percentage, of content data between any other data servers dis-
35 posed between the low end and the high end data servers, allocating for each succeeding data server more
remote from the data accessing station a greater percentage of content data for storage than an allocation for
any data server less remotely disposed from the data accessing station and a lesser percentage of content
data than the next succeeding server more remotely disposed from the data accessing station.

In a third aspect, the invention provides a data communication system for transmitting digital data between
40 remotely located system stations where each station is adapted to participate in a process of storing, account-
ing for and manufacturing media with digitized content data packaged for playback, comprising: a point of ac-
cess station for identifying selected digitized content data and for generating a communication requesting said
content data; a content data storage station having digitized content data remotely disposed from the point of
access station; an accounting station for recording transactional data for each customer point of access station
45 communication requesting digitized content data from the content data storage station; a manufacturing station
remotely disposed from said content data storage station for generating media incorporating the digitized con-
tent data requested by the point of access station from the content data storage station; and a communication
network for interconnecting the remotely disposed content data storage station with the point of access station,
the accounting station and the manufacturing station to transfer digitized content data in real-time between
50 the content data storage station and the manufacturing station.

In a fourth aspect, the invention provides a multimedia preview station including an audio system, a
touchscreen monitor and a controller for individually presenting a plurality of screens on the monitor for touch
control by a consumer to present to the consumer consumer identified and requested screen presentations,
comprising; a library screen identifying a plurality of categories of content data for screen presentation on the
55 monitor including a touch point for activating the presentation of a category; an alphabetical listing screen iden-
tifying content data for presentation by alphabetical listing of the category subject matter of the library screen
including a touch point for activation by the consumer; a search screen including monitor touch points for iden-
tifying a consumer request either alphabetically or numerically; a preview screen including monitor touch points

EP 0 649 121 A2

for previewing consumer identified audio and video listed content data transferred to the preview screen in
real-time and monitor touch points for selecting for purchase listed content data; and a purchase selection
screen including purchase selections made by the consumer, listed by item selected, quantity, format and price,
including monitor touch points for altering the item selection.

In a fifth aspect, the invention provides an accounting system including a touchscreen monitor and a con-
troller for individually presenting a plurality of screens on the monitor for touch control by an operator to present
to the operator, accounting worktools for analysis of remotely stored digitized data of consumer purchases as
a function of sales operation, quantity, monetary amount, purchase format, purchase subject matter and con-
sumer demographics, comprising: a report screen identifying a plurality of analysis reports for screen presen-
tation on the monitor including a touchpoint for activating the presentation of an analysis report as a function
of consumer sales, consumer demographics, sales unit, and sales group; a digitized data storage unit remotely
disposed from the report screen for transferring at a high speed constant bit rate the digitized data correspond-
ing to a touchpoint activated analysis report on the report screen of the touchscreen monitor; and an analysis
report screen presenting in quantified format the remotely stored digital data corresponding to consumer pur-
chases.

In a sixth aspect, the invention provides a manufacturing system for manufacturing media with digital data
content including a plurality of data servers storing digital data content remotely disposed to each other and
connected in hierarchal levels comprising: top level servers each storing a high percentage of the stored digital
data content; intermediate level servers each storing a lesser percentage of stored digital data content than
the top level servers; a high level communication network for interconnecting the top level servers and the
intermediate level servers and for transferring digital data content between the top level servers and the in-
termediate level servers in asynchronous transfer mode; end site servers each storing a lesser percentage of
stored digital data content than the intermediate level servers; a low level communication network for inter-
connecting the intermediate level servers and the end site servers and for transferring digital data content be-
tween the intermediate level servers and the end site servers in asynchronous transfer mode; and a manu-
facturing device interconnected to one of the end site servers for receiving in asynchronous transfer mode dig-
ital data content for manufacturing media with the received digital data content.

The present invention enables the provision of a digital data on-demand turnkey system wherein N number
of servers provide for 100% of content distribution of remotely stored digitized information, which information
may be previewed or reviewed in real-time, and products incorporating selected digitized information can be
manufactured within a short response time to a customer's request at a point of sale location. This technique
is carried out without intermediate buffering or caching which minimizes costs particularly at those locations
being served. The Oxford Dictionary of Computing defines a real-time system as any system in which the time
at which output is produced is significant It states that this is usually because the input corresponds to some
movement in the physical world, and the output has to relate to that same movement. It notes that the lag from
input time to output time must be sufficiently small for acceptable timeliness.

It is preferred for consumers in a retail environment, whether they be at a point-of-sale or at another cus-
tomer premise, to be able to exhaustively search and preview the content database using graphics- based touch
screens at consumer interface terminals such as a kiosk. For a proposed audio application, previews including
audio and video segments are available. Through the use of prompting screens customers are able to make
purchasing decisions by stipulating content which is available from any number of categories of subject matter
such as music types. Content can be identified by album, artist or remembered portions of material within the
content of the totality of the work.

The selected media for the manufacture and production of digital and analog software may be from a myri-
ad of different selections and can be accomplished on digital media such as CD's, analog technology such as
cassette tapes, CD ROM technology, reel-to-reel tapes, game cartridges, video disks of varying sizes, sheet
music, floppy disks and can include content which is audio, such as voice or music, video, interactive game
program and cartography, just to mention a few.

Since the system is adaptable to having an unlimited number of servers, strategic and economic network-
ing can be accomplished to allocate proportionally the preview material as well as the total content of a specific
work at the respective servers depending, in particular, on the popularity and the costs. A master server will
be situated geographically so as to be accessible to chained and networked subservers. Its geographic and
networked location is dependent upon communication network systems and subsystems costs and availability
in order to best serve a customer's premise, whether it be a retail store or similar point-of-sale or other end-
user location. The strategy of allocating between the servers the content for preview and for the final manu-
facturing process for a particular work, is strategized In order to eliminate the need for any in-store inventory
of an artist's composition regardless of format. In essence, every album in every particular label's catalog will
be available in any format and the retailer will never have a condition of being out of stock. The retailer will be

EP 0 649 121 A2

able to utilize floor space to maximize merchandising and need not feel the burden of inventory drag.

Thus , merchandising is greatly enhanced through the use of multi-media kiosks or booths individualized
to the customer and/or categories of buyers. This makes the selection, preview and purchase more personal
to the customer. Not only is the customer able to preview any number of one of a selected categories of music

5 types or whatever the content subject matter may be, including electronic games, software programs, video,
audio and images, the customer can be introduced to other albums or works of a particular artist or a particular
type of music. With the multi-media kiosk experience, the retailer has an opportunity to further enhance the
purchaser's interest in matters which possibly are collateral to the original intent, interest and desire of a cus-
tomer. Obviously, this is of great advantage to the retailer and can result in additional sales. The gathering and

10 further use of artificial Intelligence of the sort which is specific to a buyer or to a category of buyers can max-
imize the retailer's return through the use of such a system.

Both the retailer and the label companies are and have been specially interested in the ability to protect
the myriad of proprietary rights which the artists, the labels and others have through copyright protection and/or
licenses. The present inventive system provides for security mechanisms that require centralized database

is authorizations prior to the transmission of content and/or the manufacture of any of the products. In addition,
the system simplifies the financial and accounting aspects and will reduce, if not totally eliminate, pilferage
of items within the store at the point of sale.

Further, before a purchase is actually made, numerous communications are made to the potential purchas-
er ensuring that the purchaser has the properly identified items of interest and is interested in their procure-

20 ment. These aspects and the further aspects of the preview opportunities greatly reduce and minimize return
of purchases by the customer.

The total authorization, accounting and management reporting and marketing research system collects
and maintains data useful to the retailer and the label companies. With proper analysis all stations along the
merchandising trail can benefit from this information and exploit the economic opportunities available from

25 the analysis of such data.

After review, selection and proper authorization procedures have transpired, all of which are done in a short
period of time having benefitted from the expedited communication and transfer of digitized information to pro-
vide for real time handling of the customer's needs at the point of sale, the manufacturing and production proc-
ess of the media in the selected format is accomplished. Not only is the manufacturing subsystem capable of

30 generating an exact replica of the master of the selected piece or work, but the manufacturing subsystem is
also capable of producing the art work or labels and any included documentation with regard to the purchased
item. The digitized information to be replicated will be communicated to the manufacturing subsystem at the
customer premise from any one or a number of the data servers, including the master server within the geo-
graphical network of servers. As stated before, the specific server from which the information or a portion of

35 the information is to be forthcoming is dependent upon such factors as the popularity of the work, the total
length of the work, the cost of transporting the information from the server to the customer premise and other
factors which are cost effective and minimize the distance to the customer premise while maintaining the re-
quired level of response for real-time transactions.

In the above mentioned way, the present invention provides an effective means for accessing remotely

40 stored information at strategically located server locations adapted to communicate and transmit content in-
formation to a customer premise in real time and in such a manner so as to be able to manufacture media in-
corporating the transmitted information, while maintaining the integrity and providing accountability of the com-
municated and networked information to provide on-demand and real time customer satisfaction.

45 BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a pictorial representation of a system architecture including a master site having communication
with a retail store according to the principles of the present invention;

FIG. 2 is a high-level communication network structure, exemplary of a communications system for im-
50 plementing the system and method of the present invention;

FIG. 3 is an exemplary remote serving topology of a hierarchical arrangement of the servers with allocated

content represented graphically opposite the corresponding hierarchical layer of the servers;

FIG. 4 is an exemplary regional data server illustrating a uniprocessor architecture;

FIG. 5 is an alternative exemplary data server illustrating a multi- processor architecture;
55 FIG. 6 is an exemplary server architecture illustrating communication interconnecting through a ring or

bus cluster;

FIG. 7 is an alternative server architecture illustrating communication interconnecting through a switch
cluster arrangement;

EP 0 649 121 A2

FIG. 8 is an exemplary switch for connecting different communication service consumers;
FIG. 9 is an exemplary hub or router for connecting different communication service consumers;
FIG. 10 is an alternative communication connecting system including an adapter and a computer for con-
necting different communication service consumers;
5 FIG. 11 is a exemplary block diagram of an on-demand system with illustrative switching devices for net-

working between regional and central servers, according to the principles of the present invention;
FIG. 12 is an exemplary partial block diagram of a retail store system with expanded illustration of the man-
ufacturing subsystem.

FIG. 13 is an exemplary block diagram of a retail store system according the principles of the present in-
10 vent ion with the data server and other subsystems within a local area network (LAN);

FIG. 14 is a functional block diagram of the on-demand data system according to the principles of the pres-
ent invention;

FIG. 15 is a logic diagram of the on-demand data system illustrating its five major subsystems without re-
gard to local or remote location;
is FIG. 16 is an exemplary illustration of the multimedia preview station of the on-demand system according

to the principles of the present invention;

FIG. 17 is an exemplary illustration of the multimedia preview station alternatively showing a set-top video
decoder application;

FIG. 1 8 is a monitor screen illustration of the library content of the on-demand audio and video system at
20 the preview station;

FIG. 19 is a multimedia preview station screen presentation of the audio and video music categories at
the preview station;

FIG. 20 is a preview station screen presentation of an exemplary listing of artists and groups within a par-
ticular music category;

25 FIG. 21 is a preview station screen illustration of a keyboard layout for manipulation by touch for identifying

through description artist's works;

FIG. 22 is an exemplary preview station screen listing of a particular search result;
FIG. 23 is an exemplary preview station screen listing of the album works of a particular artist or group;
FIG. 24 is an exemplary preview station screen listing of selections for preview of a particular album;
30 FIG. 25 is an exemplary preview station screen depicting a single selection for audio preview;

FIG. 26 is an exemplary video preview station screen with motion video of a selection;
FIG. 27 is an exemplary preview station screen listing and itemization of selected items in the shopping
cart of the previewer;

FIG. 28 is an exemplary preview station screen presenting a previewer with the opportunity to submit or
35 cancel an order;

FIG. 29 is a screen illustration of the business analysis tool workstation illustrating the categories for de-
tailed accounting and finance information;

FIG. 30 is an exemplary screen illustration of an accounting for a particular point-of-sale outlet by product
format;

40 FIG. 31 is an exemplary screen illustration of the high volume album selections for a particular time period.

FIG. 32 is an exemplary screen view of an alternative accounting tool illustrating sales activity by album
and music category;

FIG. 33 is an exemplary graphical screen presentation of retail activity by type of music;
FIG. 34 is an exemplary graphical screen presentation of retail activity by type of media;
45 FIG. 35 is an exemplary accounting workstation screen chart showing demographic statistics for sales of

product types;

FIG. 36 is an exemplary screen presentation of accounting workstation analysis information which is store-
hours time dependent;

FIG. 37 is an exemplary central host accounting workstation screen analysis tool for local, regional and
so chain accounting information; and

FIG. 38 is an exemplary hardcopy printout report of a central host analysis accounting tool showing sales
for the regional central host by music category.

DESCRIPTION OF THE ILLUSTRATIVE EMBODIMENTS

White the present invention will be described more fully hereinafter with reference to the accompanying
drawings, in which illustrative embodiments of the present invention are shown, it is to be understood at the
outset of the description which follows that persons of skill in the appropriate arts may modify the invention

EP 0 649 121 A2

here described while still achieving the favorable results of this invention. Accordingly, the description which
follows is to be understood as being a broad, teaching disclosure directed to persons of skill in the appropriate
arts, and not as limiting upon the present invention.

Many of the deficiencies, inadequacies, restrictions, limitations both technically and commercially and the
cost prohibitions of the presently available on-demand packaged digital data for playback systems are over-
come by the present invention. Presented is a concept where a device for manufacturing media with content
data has virtual connection to a host content data machine where all message packets arrive and in sequence
to a host content data machine at a high speed constant bit rate, providing for data to be streamed from a Master
Site 102 to a Manufacturing subsystem 104 as shown in FIG. 1. In the context of the present invention a data
stream is a continuous stream of data elements being transmitted, or Intended for transmission, in character
or binary-digit form. In a preferred embodiment of such a system a data preview station 106 is also provided
and high speed data streaming is a desirable feature of the operation for preview. While K is not necessary to
have high speed data streaming for the transaction accounting operations of the Master Site 102 and a Ac-
counting Subsystem 108 at, e.g., the retail store 110, that also is preferred.

In a preferred manufacturing system with accountability, high speed data streaming would be from end
station to end station. One end station would be for data preview at booths 106. Another end station is the
manufacturing device or subsystem 104. Yet another end station is the manager's or the accounting work sta-
tion 108. While a further end station is the data content host or the Master Site 102.

An important aspect of streaming data between the end stations is for the data to be streamed so that it
can be presented so as to reproduce whatever the recorded information is at the proper reproduction rate so
that the individual using it or looking at it or listening to it can not distinguish whether or not the recorded version
is being played back from a local media, such as a video tape or compact disc, or if the individual is witnessing
it being played back remotely and transmitted through a network. The bits of information have to be delivered
to the preview or manufacturing device at a rate required by that device to recreate faithful reproduction of the
25 original recording. ,

A communications cloud, such as cloud 112 in FIG. 1 generically represents a network where included
state-of-the art topology provides for routing and transporting of data. In general, the interface into the cloud
is specified by the myriad of different network topologies. So regardless of the specific topology or type of a
backbone or switching techniques within the cloud, the data gets transported from one point to another point
30 over a wide geographical area.

The Tn designations of FIG. 1 , or T Carrier transmission systems are represented by DSn or Digital Signal
levels which make up the standard digital transmission hierarchy developed by Bell Labs. DSO represents a
single digital voice channel at 64 Kilobits per second. T 1 represents 24 DSO channels in the US., and in Europe
the designation is E 1 representing 30 DSO channels. The different worldwide digital hierarchy's are referenced
35 in their applicable standards body's such as Bellcore Technical References, International Telecommunications
Union (CCITT) and International Organization for Standardization. The OC-N designation represents the Sonet
Optical Carrier Speed Hierarchy where OC-3 is equivalent to 84 T 1 or 2016 DSO voice channels. The text
Data Network Design authored by Darren L. Spohn copyrighted 1993 by publisher McGraw-Hill, Inc. presents
design information relied upon in the development of the present invention.
40 In the illustration of FIG. 1 , coming into the customer premise site or Retail Store 110, as for example, either

a T1 or DS1 , which is 1 .5 megabits or a T3 or DS3 level which is 44 megabits per second or an asynchronous
transfer mode (ATM) which would be carried over in OC-3 optical fiber which provides 155 megabit fiber con-
nection or a combination is employed. ATM is a cell relay type of protocol which is being widely developed
throughout the telephony industry. At the other side of the exemplary cloud 112 where a serving site exists
45 and, in the illustration of FIG. 1 , Master Site 102, there is a exemplary connect to T3, DS3 which follows with
ATM running over either OC-3 fiber at 155 megabit or even up to OC-12 which is 622 megabits.

At the top level server site, the actual bandwidth or bit rate required would be dependent on the number
of, and the size and the quantity of the data that is served from that site. Multiple T3s or multiple OC-3s or
OC-12s are connected and multiplexed out to the different devices in the Master Site 102. This is also the prac-
50 tice for any of the server sites, such as Regional Site 114 and Local Site 116, as well as the top level sites or
Master Site 102.

FIGS. 8, 9, 10 illustrated three approaches to multiplex and demultiplex the aggregate bandwidth required
by the various terminals and devices contained within the customer premises to a single high speed trunk en-
tering the network. This approach is also applied at the server sites to aggregate bandwidth entering the net-
55 work. Another important function of the switches 118 or 120 in FIG. 1, and in FIGS. 8, 9, 10, is to multiplex or
aggregate multiple trunks from the network provider to the customer premise or server site.

The ATM communication technique presents a standard 53 octet or 53 byte cell which has included within
it a header comprising the virtual path identifier, the generic flow control , the virtual channel identifier, reserve

EP 0 649 121 A2

bits to be determined, a paytoad type, a call lost priority and a header error control field. A cell includes the
header of 5 bytes and a 48 byte data content payload. It is possible to provide in one of these cells of 53 bytes,
voice, video and imaging data. This cell would move at a constant bit rate speed from end station to end station
and when, for example, it reached the manufacturing device the header information is discarded and the data
5 portion of the packet is copied to the media. This media, as for example, can be CD or magnetic media for
reproduction and manufacturing at this station.

Each of these cells is assigned a path by the communications network dependent upon the network traffic
and load. In the communications industry this is known as fast packet switching. It provides an expeditious
way to move a large volume of data.
10 The feature of ATM which provides for all packets or cells being of the same length presents a condition
where small messages which are time dependent do not get held up by large messages which are not time
dependent and can be delayed without degradation of the message's value. This is because all messages are
comprised of the same 53 octet length. As hereinbefore mentioned the user has the advantage of different
levels of service depending on the user's requirements for streaming the applications data. The header on each
is of the 53 byte uniform length cells provides routing information for each of the cells in the network. In a shared
network environment, such as is present in much of today's communication networks there is uncertainty and
variability. The ATM process provides for standardization in the actual transport method or transmission meth-
od through the use of fixed sized cells, plus priority routing of time critical data as specified in the cell header.
A further feature of the present invention is made evident by FIG.1, where at the master site 102, the con-
20 cept of separating the different logical functions that are a part of the serving site into different or separate
types of machines is presented. The idea of separating the functions allows for the scaling of different sizes
of serving machines, and the specializing of the hardware configurations and even the software configuration
so that they are able to perform the identified specialized functions. The machines include the network man-
agement unit 122, the directory services unit 124, the data server unit 126, the customer data base unit 128,
25 incorporating information individualized to the customer's purchasing habits and interests and the royalty ac-
counting unit 130. The network management unit 122 checks the status of the operation of communication
links with the customer premise 110 and units within the customer premise 110. The directory for services 124
provides address functions for appropriate routing of messages requesting, for example, content data at the
data server 126 or customer data in the customer database 128.
30 An example, illustrating the separation of function issue, is applicable to the royalty accounting unit 130

where if there is a suitable application available, for example, on an IBM AS/400 system that does credit card
transactions or accounting transactions, it is desirable to adopt that kind of a machine and based on if s other
characteristics, e.g., security, adapt it for use in the banking industry for use as a machine to track the royalty
transactions that would flow in from the various customer premises. Similarly, a separate machine with spe-
35 cialized software can accomplish the network management of unit 122 at the serving site.

In addition, the data server 126 or servers can be specialized in machine type, as to the amount of storage
etc. that they manage according to the type of content. So. a system can include one type of machine that is
well suited to serve video data, and a second type of data serving machine that might be more suitable for
such data as electronic game content data or audio data. Therefore at the different serving sites, the different
40 functions are separated to run on separate machines, interconnecting them through switches or hubs or a suit-
able type of interconnecting arrangement at the serving site, and then manage the message protocols to route
the appropriate message traffic coming in from the cloud 112 to the appropriate functional processing machine.

Also in FIG.1 in the description, the local site 116 and regional site 114 are not pictorially displayed like
the master site 1 02. The bulleted list identifies some of the subset of the functions that are in the discussion
45 breakdown of the master site 102. Certain of these functions, for example, those of customer data base unit
128, might not be of sufficiently high enough volume or traffic to justify replicating the customer data base
machine 128 at ail the different serving sites. So, by functionally separating the types of functions that are
being served out of the sites, they can be distributed differently throughout the serving hierarchy. Therefore
at some number of the serving sites or maybe only at certain levels of the serving hierarchy, some of these
so functions are present while at other levels they are not.

The lowest level of the data servers or ESS servers 306 in FIG. 3 are most important in terms of the high
volume transactions, and replication of those Is done in a more generous manner. By separating the different
serving functions the opportunity is provided to choose the appropriate configuration for each machine and
use the communication switch interconnections to route the data traffic to the appropriate machine or function.
55 FIG. 2 is an example of how the exemplary cloud 112 of FIG. 1 might physically connect three different

serving sites together. In FIG. 2 there are three different service providers i.e., solid lines 202, dashed lines
204 and dotted and dashed lines 206 carriers or telephone operating companies shown by the different types
of connecting lines. In FIG. 2 there are central offices (COs) 208. where the customer premises (or the stores);

EP 0 649 121 A2

are connected. Then there are trunk connections going from central offices 208 to other central offices 208
which connect to the different server sites, i.e., Master Site 210, and Regional Site 212 and Regional Site 214.
Each of the server sites 210 and 212 has more than one set of trunks going to different service provider central
offices (COs). In the serving sites 210, 212 or 214, the included switches are used to terminate the different
5 trunks coming from the different central offices. On the back side of the switch is where the serving machines
are connected. The switching function at the server site allows the data traffic routing between the various
clouds.

In FIG. 2 there is represented three different sets of service providers 202, 204 and 206 that provide cloud
function for interconnecting not only server sites, but also the retail stores.

10 Using ATM as the interconnect with the data server, gives the physical appearance of having the entire

data content on every one of the local servers. It is still desirable to replicate some portion of that storage in
each of the local servers for transaction and frequency of requestors demand reasons, but transparently, by
communications through ATM, there is the appearance of having 100% of the data content of every catalog
listing or item at the serving site. The number of the replication sites can be regulated or reduced depending

15 upon how widespread the presence of ATM is.

The only reason to replicate the databank is to be able to spread the high demand and bandwidth amongst
all the various sites. This practice is set forth in FIG. 2. Primary servers can function as backup servers. Such,
as for example, is possible with regard to server 21 2 providing backup for server 214 when serving the central
office 208 connected between them and illustrated in FIG. 2 as the bottom and centrally most disposed central

20 ' office. The backup strategy provides for efficiency and also makes it easier to recover in the onset of break-
downs or disasters.

Being able to give the appearance of having 1 00% of the data content locally, camouflages having servers
physically disbursed throughout a geographic area regardless of size. It is evident that through implementation
of this inventive concept, directory service control function can control all the different Customer Premise sites

25 to decide from where to access and obtain the content data. This control structure still relies on having all the
data server sites connected by an ATM backbone to accomplish the necessary network.

FIG. 3 illustrates the communication network layout with emphasis on (a) a multiple-level structure of data
servers providing improved reliability and availability of data, (b) redundant connections among the servers
and their clients through multiple communication paths, and (c) distribution of on-line data according to demand

30 and on-line storage and transmission costs.

In this FIG. 3 there are several repositories each of which typically contains a portion of the totality of the
content material available. Each exemplary level in this structure is comprised of server class machines which
are large configuration systems with large amounts of on-line DASD storage plus extensive communications
capability. In the FIG. 3 example, the top most server level can contain 100% of the data which is available.

35 The top level servers sometimes called master servers or golden masters are usually provided with 100% of
the content data at very little additional cost above the 90% or other high percentage they would otherwise
store as a result of allocation strategies where the content data allocated to servers at the various levels is
based upon the frequency of content data use and is a function of the popularity of segments of the content
data and is a result of data studies from consumer demand curves. A very high percentage also provides for

40 redundancy and back up if service to a server should fail. Set forth for discussion purposes only, are arbitrary
percentages of data divided between the multiple levels of servers. In practice the percentages are variable
and are, for example, a function of such things as consumer demographics and time of year, just to mention
a couple.

The servers at this level (a), are the Top Level Servers (TLS) 302. There are several of these servers,
45 geographically placed according to several factors. The factors include: communication line availability and
communication line cost, aggregate demand from the next lower level, and other considerations such as geo-
graphic and safety/security of the data.

These servers 302 have a very high percentage of the content data, sometimes 100%, but in the case of FIG.
3 it is 92% which satisfies the least demand or approximately 35% of the content data requests.

so Below the Top Level Servers 302 in this structure at level (b) is a collection of servers 304 which contain

a considerably lesser portion of the total data (typically around 5%). Yet, this level server 304 satisfies up to
the next 20% of the demand on average, due to the popularity of the content data stored at the (b) level. These
servers are the Intermediate Level Servers (ILS) 304.

Finally, in the example of FIG. 3 there is a set of servers 306 at a level (c) which support the customer

55 premises directly. In the structure, these servers 306 satisfy the largest percent of the requests for the most
frequently selected content data. These servers are the End-Site Servers (ESS) 306. As compared to the TLS
and ILS, the ESS servers hold the minimal amount of the content material available, which is typically around
3%. However, as can be seen from FIG. 3, these servers 306 satisfy the bulk of the customer's requests for

EP 0 649 121 A2

content data, which is about 45%.

A preferred allocation of content data amongst a plurality of severs connected in hierachical levels includes
allocating <10% of the content data at the low end data server, and allocating >80% of the content data at the
data server at the high end of the hierarchy most remote from the data accessing station.

s The cost analysis is driven by an effort to achieve replicating a small percentages of physical storage at

the lower levels of the structure, so that a considerable amount of the transaction demand is satisfied while
spreading out the balance among different server sites and machines. The demand requests for the data must
be collected and an analysis made of the transaction patterns. The study results are used to minimize the data
replication and maximize (i.e.,reduce) the response time of the transaction, and distribute the data for storage

w among large numbers of servers and sites without experiencing excessive costs to replicate these amounts
of storage which are widely distributed.

For a typical request, when a data file and its content is needed and the service of a higher level server
is required a request is sent from the Customer Premise (CP) 308, to the ESS 306 which is its preferred route.
The communications network may determine that the requested ESS 306 is available directly (i.e., in a single

is path) or it may select an alternate path for this conversation. An alternate path could be selected through any
path with sufficient level of service, particularly with regard to bandwidth. In this example, alternate paths could
be found through any of the communication clouds 310, 312, or 314. Alternatively, when the message arrives
at the selected ESS 306, the data server on that site could determine that (i) it does not have the data requested
and so it could re-route the request to an ILS 304 itself or (ii) that it has too many transactions in process, and

20 routes the request to another ESS 306 which could for example be in another geographical time zone.

FIG. 4 is an exemplary uniprocessor system data server just like a standard personal computer IBM PS/2
Model 95 with a single processor 402 connected through a fast bus 406 to some amount of memory 404, a
redundant array of independent disks (RAID) 416, SCSI buss 408 to hang storage units 412 on and connecting
line 410, then some sort of I/O 414 which represents things like communications adapters.

25 FIG. 5 illustrates a multi-processor data server system where there is more than one processor 502 sitting
on a single fast bus 504 or interprocessor memory bus of some sort with multiple I/O buses 506, 508 and 510
hanging off of the system. There are several different types of machines that represent this kind of structure
ranging from the IBM 3090 types of mainframes to systems like Tricord, Sequent and Pyramid Models. There
are many kinds of multiprocessor shared memory types of machines.

30 FIGS. 6 and 7 illustrate multiple processing elements or servers 602, such as those illustrated in FIGS.

4 and 5 that are interconnected with different possible high-speed interconnects. The first example of inter-
connects, as set forth in FIG. 6, is LAN technology interconnecting various processing elements with their own
private memories and I/O devices or servers 602. This is also illustrative of a fiber channel standard or a pro-
prietary interconnect schemes as shown by way of the oval interconnect 604.

35 FIG. 7 is another distribution topology showing that using an ATM or other high speed switch 702 as an

interconnect method between those different types of processing elements or servers 602. In shared memory
multiprocessors, blocking effects and interference, from trying to access the same memory, are experienced.
Bottlenecks on the bus or in the memory are the more adverse conditions which can arise and are adverse
to the speed of the processors and the I/O devices. However, shared memory does simplify some of the soft-

40 ware structures. Since all the data is accessible from the various and different processing elements in the dis-
tributed function structures, there is less interference, because there is more message-based systems which
allows them to run asynchronously. But under such circumstances, latency or the time interval in terms of re-
questing the data for transfer and getting it transferred across whatever the interconnect scheme is, is of prime
concern. The physical topology is important but it usually results in the need for software which is more com-

45 ptex. There is a need to acquire the right message constructs with low latency, so that the form of throughput
is acceptable. The other benefit is that structures of the type in FIGS. 6 and 7 tend to scale up from small
configurations to the large and super large configurations with less difficulty than the shared memory config-
urations. In shared memory sooner or later bus limitations or locking interferences are presented which are
difficult to overcome. So the preference is a methodology that permits scale-up without undue complexity or

so waste.

Another flexibility with such structures is the ability to specialize the different types of processors, so that
a special assist for doing communications and a special assist for doing the data storage retrieval can be cre-
ated. Also specializing the different processing elements can be done, as opposed to using a single homoge-
nous type of processing element which is what is typically done in the shared memory multiprocessor. For the
55 loosely coupled structure of FIG. 7, each of these elements can also be a tightly coupled shared memory proc-
essing element itself.

FIGS. 8, 9 and 10 relate to switching, routing and other adapter and computer techniques with respect to
the service providers as herein before presented regarding FIGS. 1 and 2. These switching techniques provide

EP 0 649 121 A2

the means for gathering the incoming and outgoing messages. The incoming messages are included within a
trunk, which is a high speed aggregated collection of bandwidth. The speed is higher than the consuming side
of the system and its total bandwidth is for all the end terminals that are in the store or customer premise. The
outgoing messages are essentially the reverse process and which exit into the network in a cloud such as cloud
5 112 of FIG. 1.

The trunk in the case of FIG. 8 ranges in speed from a T1 to a T3, or an ATM and operates over fiber type
of physical connectivity. If the switch 802 is the connection media, then messages are switched from the trunk
to any one of several other communication ports. The advantage of a switch is that typically the switch 802 is
designed for a large aggregate amount of bandwidth which allows a large number of simultaneous connections

w to occur, including on up to the bandwidth limit. A less expensive alternative to FIG. 8 Is the configuration of
FIG. 1 0 which shows a computer 1 002 acting as a switching mechanism or a gateway where the trunk is coming
in through a trunk adapter 1 004 plugged in on the computer bus. In this example, it could be either a T1 adapter,
a T3 adapter or an OC-3 fiber adapter plugged in on the computer bus with other adapter cards plugged into
the computer bus, that, for example, could be LAN types of adapters. The computer 1002, with its software

15 and hardware perform a multiplexing, a routing, or a gateway type function from the LAN conversations to the
trunk adapter

FIG. 9, is an example of specialized hardware built to represent the same kind of functionality. There are
several manufacturers providing products identified as either hubs or routers that perform similar functions,
where there is one type of physical connectivity that is routed or bridged to LANs.

20 All three of these figures, FIGS. 8, 9 and 10, represent the exemplary arrangement of an incoming high
speed trunk 1 1 8 in FIG. 1 into the customer premise. The switching function occurs whether at an actual switch
or hub router or even a computer acting as a router, to take the traffic coming in from the outside network and
then switch it to the appropriate data consumer and or terminals, inside of the customer premise.

Typical switches are made out of what is referred to as switching fabric and typically are of a cross point

25 switching mechanism. The actual physical medium for the higher speed, transmission may, be on copper or
on fiber. Although there are other types of fabric that are used to perform switching functions, they usually
provide some sort of a simultaneous connection between an incoming port and an outgoing port. Crosspoint
switches are a space division form of multiplexing technique. Either the hub router or the computer gate-
way/router technique, are typically time division type of multiplexing mechanisms.

30 In the case of FIG. 1 0, the computer could be a personal computer, a workstation, or almost any computer
that has the normal input/output bus concept with I/O adapters plugged into it. In such a case the time division
multiplexing aspect of the computer bus to switch data from one adapter to another is employed. The hub and
the router are specially packaged versions of the computer of FIG. 10. Alternatively, they can include a switch-
ing fabric, i.e., a space division type of fabric. There are three types of multiplexing, i.e., frequency division,

35 space division or time division.

One of the distinctions of FIG. 11 is the cloud 1102 between the regional server site 1106 and the retail
store site 1108. The service provider can be any from the group including a local phone company, operating
company, an interexchange carrier, or a bypass company. Examples of the communication media in clouds
1102 and 1104 are copper wire, under sea cables, fiber, microwaves and satellite links.

40 Typically the T carriers or the DS, digital signal hierarchies are carried over microwaves, as well- as satel-

lites. Presently some satellite companies are exploring how to transport the higher speed carriers like the OC-
3 types of speeds across their carriers so that ATM can flow at higher speeds than 44 megabits per second.

FIG. 12 details the manufacturing and decompression aspects of the real-time transfer of data to the man-
ufacturing subsystem 104 from the repository. Here the aggregate data coming in from the cloud 1102 could

45 be ATM at 155 megabits per second, but the data going out to each one of the manufacturing machines 1202,
1204, 1206 and 1208 depends on their individual requirements. In this specific unit 118, i.e., an ATM switch,
it could in fact, be the PC that is used to de-multiplex the data and route it to each one of the manufacturing
machines at their different rates. Data comes in at real-time and is transmitted at whatever the requirement
is for each of the machines 1 202, 1 204, 1 206 and 1 208, respectively. This is accomplished whether it is at 1 .4

so megabytes per second or 1 .2 megabytes or if it was 700KB per second. The printer 1206 has a lesser need
for real-time data flow and the data transmission can be at a specific data signalling rate.

In the preferred embodiment on-demand content data system there are a multitude of different data rate
values for the various different medias, techniques, compression algorithms of units 1210, 1212, 1214 and
1216 and reproduction machines. They are variable, and dependent on the streaming requirements of the re-

55 production machine, a function of whether it is audio, or video and audio, or if it is video compression by com-
pression technique.

The preview station has a different required data flow rate then the CD writer. A CD writer, for example,
can write six times normal playback speed. A high speed data rate of 1 . 1 megabytes per second is necessary

EP 0 649 121 A2

for a six times CD writer, 1 .4 megabytes for an eight times tape writer and 7 megabytes for 40 preview booths
playing 1.4 Mbit per second MPEG compressed video or CD quality audio.

It is desirable to encrypt certain content data to maintain a higher level of security for proprietary informa-
tion and communication channels and communication switching techniques which are readily accessible to un-
authorized or unintended Interception. According data may be encrypted for example at Data Capture unite
1 32. Data is encrypted at unit 1 32 with a random pattern of characters known as an encryption key. Each cryp-
tographic key is protected and is stored in royalty accounting unit 130. Decryption of the encoded data is ac-
complished at the units 1210, 1212 and 1214 where the decryption and decompression steps are performed
in a manner consistent with the processes of encryption and compression at the time of data capture. Units
1210, 1212 and 1214 also provide the capability of buffering to accomplish streaming to the CD writer 1202,
tapewriter 1204, color printer 1206 and alternate unit 1208.

A preview station 1303 of FIG. 12 has a need for video considerations. In some cases it will have MPEG
video compressed and alternatively RTV or PLV compressed DVI or any of the various compression formats
supported by software only motion video playback. For effective streaming, the conditions must exist for con-
tinuous transmission once commenced, until completion of the data segment, which is usually the end of the
recording or data stream. This continuous transmission must be guaranteed not to fall below the speed required
by the playback device to sustain continuous full motion video.

The communications asynchronous transfer mode (ATM) with its switching technique provides advantages
over broadband communications, and its incorporation within the present invention presents opportunities for
handling all types of traffic and is protocol insensitive because of the bandwidth and the high speed bit rate.
It is particularly suitable for voice, video, image and data transmission as identified in the present invention.

Although the operational aspects of the innovative on-demand digital information system has been dis-
cussed in part with regard to Figs. 1-12 and the corresponding communication and network features of the
present invention, Fig. 13 and its description presents a total overview of atypical preview, order, authorization,
manufacture and purchase process for the average customer. The total process has been divided into com-
ponent substeps identified as the Order Process (OP), the Authorization Process (AP), the Manufacturing
Process (MP) and the Purchase Process (PP). The integral and alternate steps of each of the OP. AP, MP and
PP processes are identified by general steps which are sequentially numbered with the lead letters a, b, c and
d, respectively. These sequenced steps identify the logic flow of a typical order from the initial browsing
through to the purchase.

Certain terms, as defined below, will aid in the understanding of the logic and flow of the order and its
subsequent handling as hereafter described for FIG. 13. The definitions include Transactional Terms. Database
Tables and Process Messages.

Transactional Terms:

Accounting Server - a set of computer programs which can and typically do serve as the central point

of control within a store. These programs provide the central data base for all or-
ders originating in a store, perform all financial accounting (related to orders
placed through this system) for a store, and perform all authorization interchanges
with the Authorization Server.

Authorization Server - a set of computer programs which makes decisions, authorizing or denying ap-
proval to manufacture a product.
These programs maintain several databases such as:

- a database of products available;

- a database of royalties owed to the company, such as a label company, owning a product; and

- a database providing criteria for determining if a given retail outlet is authorized to sell a given product.

These programs provide a decision making process for several stores. These pro-
grams typically execute at a site that is separate from a retail store.

Customer Premise - that location from which a customer can interact with the on-demand system, in-

cluding, but not limited to, such interactions, like previewing, ordering, product
pick-up and maintaining transactional records.

Device Controller - a set of computer programs which control the manufacturing process. This server

is located in a retail store.

I tem . a manufactured product or work, suitable for purchase. As an example, it is that

which a customer might procure In a traditional purchase. It includes primarily the
media on which the information is recorded plus secondary materials such as a
case, labels, pamphlets, case liner, etc. In an exemplary retail system, an item

EP 0 649 121 A2

Piece

would be a compact disc, a cassette tape or electronic video games contained in
a package along with appropriate labels and printed materials.

- a record description of a single work, in some quantity, which is to be purchased.
A Line Item consists of one or more identical Items. In addition to a count of the
number of copies desired, is an indication if compact disc or cassette tape is to
be used as the recording media.

- a collection of Items a customer wishes to purchase. Orders are made up of one
or more Line Items.

- one component part of an Item, such as:

- the pamphlet placed inside a compact disc case,

- the label placed on the case, or

- the tape cassette.

Line Item

Order

Point of Sale

Preview Station

Work

- a terminal such as a cash register, located at the point where a customer makes
a purchase. This terminal includes a computer program that allows a sales person
to view, and, if appropriate, change an order.

- a location, including devices with which an individual accesses a sample of a
work. Typically the access is to demonstrate the capabilities of the offered product
at a preview by the viewer with the prospect of leading to a sale of the product to
that individual. A station of this type would generally be located in a retail sales
environment such as a store or kiosk. Orders for product may be placed at the Pre-
view Station.

- an Item being offered for preview and/or for sale.

Database Tables:

Grant Table - a persistent record of all manufacturing transactions requested on a per store basis.

Contains one record for each I tern ordered describing the authorization status of that Item.
Line Item Table - a persistent record of all ordered Line Items for ail orders on a per store basis.
Make Table - a temporary record of all Items ordered but not yet manufactured. There is one record

for each Item describing what is to be manufactured and the stage of the manufacturing

process.

Orders Table - a persistent record of ail Orders on a per store basis.

Process Messages:

op_approved - the message sent from the Authorization Server to the Accounting Server when an

Item is approved for manufacturing.
op_cancel - the message sent from the Point of Sale station to the Accounting Server to indicate

that an order has been canceled prior to the authorization request and the order should

be removed from the system.
op_cannot_make - the message sentfrom the Device Controllerto the Accounting Serverwhen a Piece

of an Item has been unsuccessfully manufactured.
op_denied - the message sent from the Authorization Server to the Accounting Server when an

Item is not approved for manufacturing.
op_grant - the message sentfrom the Accounting Server to the Authorization Server requesting

authorization to manufacture a single Item.
op_grantJailed - tne message sent from the Accounting Server to the Point of Sale station when an

authorization request is denied by the Authorization Server.
op_made - the message sentfrom the Device Controllerto the Accounting Serverwhen a Piece

of an Item has been successfully manufactured.
op_make - the message sent from the Accounting Server to the Device Controller to indicate

what Items are to be manufactured.
op_/nake_failed - the message sent from the Accounting Server to the Point-of-Sale station when an

Item is unsuccessfully manufactured,
opok - the message sentfrom the Point-of-Sale station to the Accounting Server to indicate

that an order has been validated, and paid for, by the customer.
op_order - the message sent from the Preview Station to the Accounting Server to indicate that

an order has been entered into the system.

EP 0 649 121 A2

op_order_complete - the message sent from the Accounting Server to the Point-of-Sale when all Items in
an order have been completed either successfully or unsuccessfully.

Order Process

a01) The customer, as a prospective purchaser, enters the store, and accesses one of a multitude of mul-
timedia preview stations 1303, at a preview subsystem 1302 as shown in FIG. 13. In this example, the
preview station is the Customer Premise. The customer inserts a membership card into a reader as more
fully explained for FIGS. 1 6, 1 7 and 28 and sets about previewing various works. Through the touch screen

w panels provided as shown in part and by way of example in FIGS. 18-27, the customer can choose to pre-

view and listen to certain works, or to listen to works while viewing an accompanying video. A screen icon
of a pair of headphones 2402, as seen in FIG. 24 and another icon of a TV set 2404 are used to inform
the customer of the availability of preview materials. When the customer has made a purchase decision,
commands at touch pads 2406 or 2408, describing the format of the desired items, are entered on the touch

is screen.

a02) A Preview Station 1 303 of FIG. 1 3 prints a memorandum of the Order 1 602 for the customer, as shown
in FIG. 16.

a03) The Preview Station of the subsystem 1302 forwards a message describing this order (i.e., message
type op_order) to the Accounting Server 1304 of the accounting and management subsystem 1306.
20 a04) The Order is stored in the Orders Table database on the Accounting Server 1304. The order infor-

mation includes:

- preview station number

- date and time of day

- a Line Item Table describing the details of the purchase, including:
25 - artist

- title

- number of copies

• media: compact disc or cassette tape.
a05) The customer approaches the checkout area of the store included within the accounting and man-
30 agement subsystem 1306, and hands a personal membership card or other suitable identifying card to

the sales clerk.

a06) The sales clerk inserts the customer's membership card through a card reader at the Point-of-Sale
(POS) terminal 1308.

a07) The POS 1308 sends a remote database query message to the Accounting Server 1304 providing
35 this customer number.

a08) The Accounting Server 1304 searches the Orders Table database for this customer's Order. If the
Order cannot be located, the Accounting Server 1304 returns an error to the POS 1308. However, if the
Order is located, the Accounting Server 1304 returns a copy to the POS 1308 for display. Included is an
Order number identifier.

40 a09) Once the customer's Order has been displayed at the POS 1 308, the sales person will ask if the cus-

tomer wishes to make any changes. Changes which are possible include:

- change quantity of any Line Item;

- delete Items;

- change format of ordered Items (e.g., com pact disc to cassette tape or cassette tape to com pact disc);
45 - add new Items not previously entered;

- cancel the entire Order.

a1 0) As an alternative to the above POS purchasing sequence, the customer could simply approach a sales
person immediately upon entering the store and have the sales person enter an Order. In this case, the
procedures related to the activities at the preview station subsystem are not necessary. (The Order will
so not be found at step a08 but the sales person will be able to enter the order in its entirety at step a09.)

a11) The customer now makes payment via cash, check, or credit card, according to the accepted proce-
dure for each particular store or retail outlet.

a12) When the sales person completes collection of the payment, the POS 1308 will send an op_ok mes-
sage to the Accounting Server 1304. At this point, all the information about the Order is in the on-demand
55 digital information system. The system is now ready and will automatically proceed to authorize the Order,

to credit royalties to any party or entity so entitled as set forth in pre-existing agreements, and to issue
manufacturing orders to generate the customer's selections.

BP 0 649 121 A2

Authorization Process

b01) When the Accounting Server 1304 receives the op_ok message, it replaces the copy of the Order in
the Orders Table database. This is to account for any changes made by the sales clerk at the POS terminal
5 1308.

b02) The Accounting Server 1 304 updates the Grant Table and Make Table with one record per Item, ini-
tialized to blank.

b03) Before any manufacturing can begin, an authorization must be obtained and an Authorization proce-
dure is performed. This insures that only works authorized or approved for manufacture are manufactured

10 and that any and all royalty requirements are correctly recorded.

b04) The Accounting Server 1304 sends an op_grant message to the Authorization Server 1312 of the
Authorization Server subsystem 1310 for each Item in each Line Item. (Note: If this step is reached due
to a manufacturing failure c13 step, only the failing Items will have grant requests sent.)
b05) The Authorization Server 1312 performs various authorization checks on each op_grant message

is and returns an op_approved or op_denied message to indicate ok/not ok to manufacture that particular

Item.

b06) The criteria for denying authorization may be objective or may include subjective rules.
Exemplary criteria include:

- release date is later than today's date;

20 - a particular store's credit rating is exceeded;

- certain retail outlets may want to prohibit sale of certain works;

- a content provider's or label company's unique rules; and

- copyright license for a work does not exist in the country where the store is located.

b07) If an item is rejected for any reason, an op_denied message is returned for that Item by the Author-
25 ization Server 1312 to the Accounting Server 1304.

b08) When op_denied is received for an Item, the Accounting Server 1304:

- updates the Grant Table, Line Item Table, and Order Table reflecting what was denied; and

- the denied item is removed from the Make Table.

b09) Since more than one Order at a time can have outstanding authorization requests, if this is the last
30 Item in a particular Order, the Accounting Server 1 304 marks the Order complete and sends an op_make

message to the Device Controller 1314 in the Manufacturing Subsystem 1316.

b10) If there are more items in this Order, the Accounting Server 1 304 loops back to Authorization Process
step b07).

b11) If an op_approved message is received for a Line Item, the Accounting Server 1304 updates the Grant
35 Table, Line Item Table, Make Table, and Orders Table reflecting what was approved. It will then loop back

to Authorization Process step b09.

Manufacturing Steps

40 In the specification below, the Device Controller 1314 configuration depicted in Fig. 13 is representative of
those machines which are necessary to operate the manufacturing machines, such as the CD writer 1318 and
the tape writer 1 320 in a particular retail store. The actual number of manufacturing devices and their location
may differ from that illustrated in FIG. 13 based upon a given store's requirements.

c01) When the Device Controller 1314 receives the op_make message from the Accounting Server 1304
45 the Order contains only those Items which have been authorized and are to be manufactured.

c02) The Device Controller 1314 displays the Order as a set of manufacturing instructions on the Manu-
facturing Technician's Terminal (MTT) 1322. Each Item to be manufactured is displayed as a single line.
c03) In the case of media manufacturing, when the appropriate machine such as the CD writer 131 B or
tape writer 1320 is ready, the technician presses a FETCH icon on the MTT 1322. This sends a message
50 to the respective Device Controller 1314 for that manufacturing machine. The technician then presses a

START icon on the MTT 1322 to initiate actual manufacture.

c04) The Device Controller 1314 will then establish a dialogue with the Data Server 1324 of the Data Server
Subsystem 1 326 and begin the content data transfer to the exemplary machine 1 31 8 or 1 320 indicated in
the Item.

55 c05) For any printed material required, the MTT 1 322 sends a message to the Device Controller 1314 iden-

tifying the data for printing..

c06) The Device Controller 1314 will then establish a dialogue with the Data Server 1 324 and begin trans-
ferring the printable material data to the printer 1 328.

EP 0 649 121 A2

c07) In both manufacturing steps c04 and c06, the data being transferred may be encrypted or compressed
or both or neither. Before the data is actually routed to the manufacturing machine such as machine 1318,
1320 or 1 328, a decryption and/or decompression step is likely necessary. The cryptographic key needed
for the decryption step may be contained in the original op_approve message of Authorization Process
5 stepb11.

c08) For each Piece of an Item successfully completed, the Device Controller 1314 sends an op_made

message to the Accounting Server 1304.

c09) The Accounting Server 1304 notes each op_made message:

- when all pieces in an Item are made successfully, that item is marked complete;
10 - when all Items in a Line Item are complete, that Line item is marked complete; and

• when all Line Items in an Order are complete, that Order is marked complete,
d 0) If an error occurs at any step of the Manufacturing Process, the recovery steps depend on the failure
point:

- for a failure in the printing step, the technician must manually retry that step.
is c11) For a failure in writing a media device such as at units 1318 or 1320 that Item will be scrapped and
a new authorization sequence must be initiated.

c12) If any Piece of an Item does not complete successfully, the Device Controller 1314 sends an op_can-
not_make message to the Accounting Server 1304.

c13) The Accounting Server 1304 receives the op_cannot__make message. The order number is used to
20 retrieve the original Order from the database. If the failing Piece is media, as opposed to printed matter,

a new grant request is constructed for the failing Item and the processing resumes above at step b04 of
the Authorization Process. This step and the associated authorization grant / royalty accounting / manu-
facturing steps above step b04 through step c13 are repeated *n* times - where n is a retry count that is
stored in an operating characteristics file in the Accounting Server 1 304. When a failure has been re-
25 tried *n' times, the Accounting Server 1304 will make a final determination that this Item cannot be deliv-

ered and processing will continue at Purchase Process step d01.

c14) The original Order remains on the account for the requesting retail store, i.e., normal royalties remain
charged against that store. That store manager must subsequently process a Return Manufacturing Au-
thorization request to obtain credit for the failed Pieces. This insures that no duplicate media may be pro-
30 duced without royalty payments being accrued to the content owner.

Purchase Process

d01) Once the Accounting Server 1304 has received a response for all Pieces of all Items in the Order
35 (either op_made or op_cannot_make and the retry count is exhausted), it accesses the original Order and

processes the Items from top- to-bottom matching as many Line Items requested to those completed .

d02) Any Items whose retry counts were exhausted [see Manufacturing step c13)] must have their selling

amounts backed out from the total charges for the Order. A credit will need to be issued, for a charge sale,

or a cash refund, for a cash sale, according to the respective store's operating policy.
40 d03) The Accounting Server 1 304 makes a final update to the Order so that it reflects what is actually being

delivered to the customer. A message is sent to the POS 1308 that the Order is complete.

d04) When the POS 1 308 receives this message, it queries the Accounting Server 1 304 for a copy of the

updated Order.

d05) The Accounting Server 1304 sends a copy of the updated Order to the POS 1308.
45 d06) The POS 1308 prints the final receipt and the sales person is informed (by a message to the screen)
and notifies the customer that their Order is ready for pickup.

FIG. 14 is a functional block diagram of the present inventive on-demand digital data system and is dis-
tinguishable from the foregoing figures as hereinafter mentioned. It identifies certain products or manufactur-
ers as exemplary sources of devices for practicing the present invention. Manufacturing controller unit 1402

50 controls the two CD writers 1318 described in FIG. 13 and identified as CD writer unit 1202 in FIG. 12 in a
manner to operate the two machines at two different speeds, i.e., twice real-time (2x) and 6 times real-time
(6x). Suitable CD writers are available from various manufacturers including the Eastman Kodak company.
Manufacturing controller unit 1404 controls the A/D interface 1 4 1 0 for the special purpose devices built for what
is tape writer 1320 of FIG. 13, as would a controller 1314 for the tape writer interface 1330 in FIG. 13. Manu-

55 facturing controller unit 1404 is also the interface for the Sun Microsystem's Workstation unit 1412 and a spe-
cific Xerox color copier or color printer 1328 of FIG. 13. FIG. 14 Illustrates the Manufacturing Control Console
1406 optionally connected to the Retail Host Server 1408.

FIG. 15 has been discussed in some considerable detail with relation to FIG. 13. Further, it shows, by way

EP 0 649 121 A2

of example only, certain illustrative programs or systems such as PARADOX which is generally hereinbefore
referred to as a database. The Presentation Manager program shows that on-demand systems can be made
to operate with the OS/2 operating system, as one such system. Likewise the booth 1502 is a multimedia pre-
view station.

5 The communication process and the flow of messages or conversation between the subsystems of a typ-

ical on-demand system was described with regard to FIG. 13. In FIG. 15 there is provided a block diagram of
a typical system encompassed by that of FIG. 13, and illustrates the presence of the communication engines
of each of the typical subsystems, i.e., the Preview Booth station 1502, the Retail Accounting station 1504,
the Central Host Accounting station 1506, the Manufacturing Control station 1508 and Point-of-Sale station

10 1510.

For the purpose of this presentation, a conversation is as a group of commands and responses that are
logically related. For example, the set of commands and responses exchanged by the Retail Accounting sub-
system 1504 and the Manufacturing Control Application 1508 is know as the 'Manufacturing Conversation*.
The five clearly defined conversations of FIG. 15 consist of bi-directional, variable-length, point-to-point mes-
15 sages. These messages can be transported by any suitable communications protocol such as NetBios, TCP/IP,
SPX/PIX, etc. Guaranteed delivery is a requirement of the communications protocol selected.

Consider the following conversations:

20 , ,

Conversation

Description

Grant

Established between the Retail Accounting 1504 and the Central Host Accounting
1506 for the purpose of requesting and granting the manufacture of a music or audio
product.

Manufacturing

Established between the Retail Accounting System 1504 and the Manufacturing Con-
trol Application 1508 to submit an order to reproduce a particular audio or video prod-
uct.

30
35

Stream

Established between the customer preview station (PS) and the file server in order to
transfer audio and video data to the PS for user interaction. This same conversation is
also established between each device controller machine and the Data Server to
transfer data to the CD and tape writers as illustrated by way of example, in more de-
tail in FIGS. 11, 12, 13 and 14.

Order

This conversation is established between the PSs and the Retail Accounting System
1504 for the purpose of placing and order put together by a customer.

Device

This conversation is established between the Manufacturing Control Application 1508
and a device server. Typical device controllers are CD writers, tape writers, and color
printers, as shown in FIG. 12. This conversation is strictly used to transfer control
command to and from the device controllers. The device controllers themselves estab-
lish a Stream conversation of their own with the Data Server.

45 In order to provide for platform and operating system independency, the unsigned long type in all of our
data structures is preferred. The unsigned long data type can be easily manipulated by DOS, OS/2, Windows,
and UNIX or X Windows, or other operating environment

The message structure is such that each message consists of a header followed by message-dependent
information. The handler of a particular message knows about the contents of the message-dependent section

so of the message. Message handlers are hereinafter discussed. The message is structured as follows:

EP 0 649 121 A2

typedef struct _msg
{

unsigned long dwOpCode;

unsigned long dwSequence

unsigned long dwFlags ;

unsigned long dwBuffLen;

unsigned char Bufferfl];

} MSG , *PMSG;

Acommon communications application program interface (API) used throughoutan exemplary on-demand
system assumes that a Netbios transport layer is available, or that a Netbios emulation can be readily imple-
mented. Such is the case for NetWare, Banyan Vines, and other network operating systems.

Further, the communications engine hides the sequencing of the messages as well as retries and re-con-
nections.

Incoming messages have the following data structures:

* The following typedef defines a pointer to a function that handles
incoming messages
*/

typedef unsigned long (*PMHFN) (unsigned long dwSessionId,FMSG pMsg);

typedef struct _rosg
{

unsigned long dwOpCode;
unsigned long dwSequence;
unsigned long dwFlags ;
unsigned long dwBuffLen;

EP 0 649 121 A2

unsigned char BufferflJ;
} MSG , *PMSG;
typedef unsigned long LISTHANDLE ;

II Error codes

#define ERROR_ALREADY — INITIALISED
#define ERR0R_MEM0RY_ALLOC
^define ERR0R__N0T_INIT1ALIZED
tfdefine ERROR_IMVALID_PTR
ttdefine ERROR_HANDLER_NOT_FOUND
#define ERROR_LANRESOURCES

#define ERROR_name_not_registered
# define error_no_more_ncbs
tdefine ERR0R_S UBMITING_NCB
fldefine ERR0R_CONNECT_IN_PR0CESS
♦define ERR0R_N0T_CONNECTED
♦define Error_connecting
♦define ERROR_SENDING
♦define ERROR_INVALID_SESSI0NID
♦define ERROR_COMH_RESOURCES
♦define ERROR_SEND_TIKEOUT
♦define error jtoo_many_cohmands

1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017

♦ifdef OS2_COMMSYS

♦define ERROR_ACCESSING_SEM

♦endif

2001

II Session flags

♦define SF_CONNECTED
# define SF.CONNECTING

♦define sf_pendreceive

♦define SF_PENDLISTEN
tdefine SF_FENDCALL

0x00000001
0x00000002
0x00000004
0x00000008
0x00000010

The following is a continuation of the description for the programming interface. More detail about the va-
rious conversations are hereinafter discussed. For example, later sections describe the conversation-depend-
ent part of the message as it Is defined for each conversation.

Initialization of the communications engine is necessary by each process that uses its services.

The global default message handler and the global exception handler MUST be passed to this function.

EP 0 649 121 A2

NULL values are not allowed. If this function fails the Comm Engine cannot be used. It is necessary to check
the return code.

PMHFN pfnDefMsgHandler , // Pointer to global def msg handler
PMHFN pfnExceptionHandler) ;// Pointer to global exception handler

Each message is handled by a message handler, that is, when a message is received the proper message
handler is invoked. If there is no message handler defined for the incoming message the Communications En-
gine invokes the default message handler for that conversation (or session) if one has been defined. If one
has not been defined the Comm Engine invokes the global default message handler which is guaranteed to
exist.

The following functions hide the implementation details of message handler support and allow the creation
and maintenance of handler lists. The first step in creating a session is to create a handler list and adding the
various message handlers to it.

int CreateMessageHandlerList( // Create an arbitrary handler list

int InitCommSys tem(
int adapters.

// Initialize the system, create chain
// Number of adapters in this machine

LIS THAN DLE *hList) ;

// Return id of arbitrary list here

int InstallMeesageHandler (
LISTHANDLE hList,
unsigned long dwOpCode,

// Add a message handler to the list

// ID of arbitrary handler list

// Opcode of handler to be installed

PMHFN pfnMessageHandler) ;

// Pointer to function handling message

int ReplaceMessageHandler (

// Replace an existing mesage handler

LISTHANDLE hList,
unsigned long dwOpCode ,
PMHFN pfnMessageHandler);

// ID of arbitrary handler list

// Opcode of handler to be replaced

// Pointer to funct ion handling message

int Regis terDef aultMessageHandler <
LISTHANDLE hList.
PMHFN pfnDefaultHandler) ;

//Install default handler for this list

// ID of arbitrary handler list

// Pointer to default handling functn

int RegisterExceptionHandler(

LISTHANDLE hList,

PMHFN pfnExceptionHandler);
function

// Install default exception handler
// ID of arbitrary handler list
// Pointer to exception handling

The following functions form the core of the Comm Engine. These functions are used to create server ses-

EP 0 649 121 A2

sions as well as to connect with remote server sessions.

int RegisterLocalName( // Register a name a-la-netbios

short iAdapter, // Network adapter number (0 through 4)

unsigned char *pszLocalName) ; // Name to be registered

int CreateServerSession< // Initiate and receive all

unsigned char *pszLocalName , // Name client connects to

unsigned char *pszRemoteNames , // Name expected to connect < n *" for

« ALLl)

LIST/HANDLE hList, // ID of handler list

unsigned long *pdwSessionId) ; // Pointer where session id returned

int ClientConnect( // client connects to server

unsigned char *pszLocalName , // Name of this client

unsigned char *pszRemoteName , If Name of the server session

LISTHANDLE hList, // ID of handler list

unsigned long *pdwSessionId) ; // Pointer where session id returned

int TerminateSession< // Ussually client terminates session

unsigned long dwSessionld) ; // Id of session to be terminated

int QuerySessionStatus( // Query the status of a given session

unsigned long dwSessionld); // Id of session to be queried

The transport manager section includes three functions. These functions allow a message to be received
and its message handler invoked, as well as sending a message across the link.

EP 0 649 121 A2

int SendNetMessage<

Client / server sends message

unsigned long dwSessionld.

Id of established session

unsigned long dwOpCode ,

Opcode to send to receiver

unsigned long dvFlags,

Value of flags to send to receiver

unsigned long dwBuffLen,

Length of buffer to send

void *pBuf fer ) ;

Pointer to buffer data to send

unsigned long QuerySessionS tatus(

Returns session status flags

unsigned long dwSessionld,

Id of established session

int DispatchNetMsg(
int bHainLoop) •
loop, FALSE otherwise

// Invokes message handlers

// Must be TRUE if called from main

The previous function deserves a special explanation. This function allows to "cooperatively" dispatch
messages to the appropriate handlers and the TRUE value is only used when D is patch NetMsg is called from
within the main loop.

For example, the following code should replace your traditional GetMsg/DispatchMsg loop in an OS/2 Pre-
sentation Manager application:

bDone = FALSE;

while {! bDone)
I

fResult = WinPeekMsg(hab, (PQMSG)&qmsg, (HWND)Nur,L,
(ULONG)NULL, (ULONG)NULL,PM_REMOVE) ;

if (fResult)
{

if (qmsg.msg == WM_QUIT)
bDone = TRUE;

else

WinDispatchMsg(hab, (pn>HSG)&qmsg) ;

}

DispatchNetHsg(TRUE) ;
]

A similar approach should be used under MS Windows.
55 If you are using the Comm Engine under PC DOS or OS/2 (non-PM application) you should have a loop

in your main as follows:

EP 0 649 121 A2

while (TRUE)
{

DispatchNetHsg(TRUE) ;

}

If processing of a particular message will take a long time (more than one second) you should call Dis-
patchNetMsg from within the handler. Make sure you call It with its parameter set to FALSE as follows:

MessageHandlerl (...)
{

for (i - 0; i < n ; i++)
{

// the following process takes a little while to execute

DispatchNetMsg(FALSE) ;
)

}

The following example uses a hypothetical sample conversation with two (2) opcodes between a server
application and a client application. This code ignores run-time errors for the most part.

For example if it is desired to start a server and a client in machine #1, and another server and another
client in machine#2 type the following:

At the OS/2 prompt in machine #1:

sessions -Ss1.c3 -Ss2.c4 -Cc1.s3 -Cc2.s4

At the OS/2 prompt in machine #2:

sessions -Ss3.c1 -Ss4.c2 -Cc3.s1 -Cc4.s2

The command-line option -S defines a server and -C defines a client- For example, -Ss1 .c3 defines a ser-
ver named s1 that accepts a connection from a remote client named c3. Option -Cc4.s2 defines a client named
c4 that connects with a remote server named s2.

EP 0 649 121 A2

ttpragma pack(l)

5 Mdefine INCL_DOSPROCESS

♦♦include <os2.h>

. ^include <stdlib.h>
♦♦include <stdio.h>
♦♦include <string.h>
♦♦include <malloc.h>
is ^include <memory.h>

^include <ctype.h>
♦♦include <time.h>

20 ^include <conrunsys.h>

int main(int argc,char **argv) ;

int DefMsgHandler(unsigned long dwSessionId # PMSG pMsg) ;

int ExceptionHandler(unsigned long dvSessionId,PMSG pMsg) ;
int MessageHandlerl (unsigned long dwSessionId,PMSG pMsg) ;
int ClientHandlerl (unsigned long dwSession!d,PMSG pMsg) ;

♦♦define

OP_REQUESTl
0P_SERVICE1

100

♦♦define

200

t main(int argc f char **argv)

int rc = 0, i, j ;
LISTHANDLE hList = 0;
L1STHANDLE hCList = 0;
char *p;

EP 0 649 121 A2

unsigned long dvServerSessions | 10 ] ;

unsigned long dwClientSessions | 10 ] ;

int nSSessions = 0;

char E2ClientName[32] ;

int nCSessions = 0;

PMSG pMsg ;

char szArg[32J;

if (argc < 3)
f

printfC'usage: SESSIONS -cClientl . Serverl -cClient2.Se
sSetverl. Client? . . . \n M ) ;

DosBeep(1000, 100) -

return(-l) ;

}

25 rc " InitCommS y ste m(l / DefMsgHandler,ExceptionHandler) ;

if (rc != SUCCESS)

printf( ,l InitialiEing Comm System (%d)\n",rc) ;
if (rc != SUCCESS)
30 return(DosBeep(1000,100) ) ;

for (i = 1; i < argc; i++)

{

35 strcpy(szArg,argv[ i]) ;

if (szArg[0] == '- ' )
{

if (toupper(szArg[l ]) == 'C )
{

p = (char *) strchr(szArg, ' . ' ) ;
if (p « 0)
{

printf ("Invalid command line argument\n\n" ) ;
return(DosBeep(1000, 100) ) ;
}

*P = 'NO';

strcpy(szClientName,&szArg[2]) ;

rc = RegisterLocalNamc(0 ( szClientName) ;
55 if (rc I* SUCCESS)

BP 0 649 121 A2

printf ("Registering %s

\n" , szClientName , rc ) ;

if (rc != SUCCESS)

return(DosBeep( 1000 ,100)) ;

}

else if (toupper(szArg[l ]) == 'S')
{

p = (char *) strchr(szArg, ' . ' ) ;
if (p == 0)
{

printf ("Invalid command line argument\n\

return(DosBeep(1000 , 100 ) ) ;

}

*P - *\0' ;

strcpy(szClientName,&szArg[2 ]) ;
*P = ' . ' r

rc = HegisterLocalName(0 / szClientName) ;
if (rc != SUCCESS)

printf ("Registering %s (%d)\n" ,&szArg[2J
if (rc != SUCCESS)

return (DosBeep( 1000 , 100)) ;

}

rc = CreateMessageHandlerList(&hList) ;
if (rc != SUCCESS)

printf ("Creating message handler list <%d) \n n , rc) ;
if (rc != SUCCESS)

return(DosBeep(1000, 100) ) ;

rc = InstallMessageHandler(hList,OP„REQUESTl,MeBsageHandlerl) ;
if (rc != SUCCESS)

printf ("Installing message handler 1 <%d)\n" , rc) ;
if (rc != SUCCESS)

return(DosBeep(1000 , 100 ) ) ;

rc = RegisterDefaultMessageHandler(hList,0) ;
if (rc != SUCCESS)

printf ("Registering default message handler 1 (%d)\n",rc);

EP 0 649 121 A2

if (rc != SUCCESS)

return(DosBeep(1000,100)) ;

rc = RegisterExceptionHandler(nList,0) ;
if (rc != SUCCESS)

10 printf ("Registering default exception handler 1 (%d)\n" ,rc) ;

if (rc != SUCCESS)

re turn (DosBeep ( 1 000 , 1 00 ) ) ;

15 II Create local server sessions

for (i - l; i < argc; i++)
{

strcpy(szArg,argv[i]) -
if (szArg[0] == 1 - ' )
f

if <toupper(szArg[l ]) == •S')
{

p = (char *) strchr(szArg, ' . ' ) ;
if (p — 0)

{

printf ("Invalid command line argument\n\n" ) ;
re turn (DosBeep( 1000 , 1 00) ) ;

}

35 *P = '\0' ;

rc = CreateServerSession(

&szArg[ 2 ] ,p+l , hLis t , &dwServe rSessions f nSSessions++ J ) ;
40 *P = ' . ' ;

if (rc != SUCCESS)

printf ("Creating %s session (result = %d) (
session

(%ld)\n",szArg, rc,dwServerSessions [nSSessions-1 J) ;

if (rc != SUCCESS)

re turn ( DosBeep ( 1000 , 1 00 ) ) ;

}

EP 0 649 121 A2

rc - CreateMessageHandlerList(&hCList);
if (rc != SUCCESS)

printf ("Creating client message handler list (%d)\n",rc) ;
if (rc SUCCESS)

return(DosBeep(1000,100) ) ;

rc = InstallMessageHandler(hCList,OP_SERVICEl f ClientHandlerl)?
if (rc != SUCCESS )

printf ("Installing client message handler 1 (%d)\n M , rc) ;
if (rc != SUCCESS)

re turn (DosBeep( 1000, 100 ) ) ;

rc = RegisterDefaultMessageHandler(hCList,0) ;
if (rc != SUCCESS)

printf ("Registering client default message handler 1

<%d)\n»,rc) ;

if (rc != SUCCESS)

r e tu rn ( DosBeep (1000,100));

rc = RegisterExceptionHand!er(hCList,0) ?
if (rc != SUCCESS)

printf ("Registering client default exception handler 1

(%d)\n",rc) ;

if (rc SUCCESS)

re turn (DosBeep (1000 ,100)) •

printf("\n\nSystem properly initialized. Press any key to connect

");

fflush(stdout) ;
getchar() ;
printf ("\n") ;

// Connect to remote servers

for (i = 1; i < argc; i++)
{

strcpy ( szArg , argvf i ] ) ;
if (szArg[0] == '-•)
{

if (toupper(szArg[l ]) == 'C')

EP 0 649 121 A2

{

p = (char *) strchr(szArg, » . • ) ;
5 if (p == 0)

{

printf ("Invalid command line argument\n\n

return(DosBeep(1000,100) ) ;

}

*P = *\0 l ;

strcpy(szClientName,&szArg(2]) ;
1S rc = ClientConnect(szClientName,p+l JiCList,

&dwClientSessions[nCSessions++] ) ;
*P = ' . ' ;

printf ("Connecting with server %s (result - %d)
20 session

(%ld)\n" ,p+l , rc,dwClientSessions[nCSessions-3 J) ;

if (rc != SUCCESS)

return(DosBeep(1000, 100)) ;

25 }

}

30 j = 0;

while (1)
i

for (i = 0; i < nCSessions; i++)
{

printf ("Sending OP_REQUESTl to session
%ld. . .\n" ,dvClientSessions[i]) ;

rc = DispatchNetMsg(l) ;
rc =

SendNe tMessage (dwClientSessions [ i ] ( 0P_REQUEST1 ,0,0,0);
if (rc != SUCCESS)
{

while (rc == ERROR_CONNECT_INJPROCESS )
{

DispatchNetMsg(l) ;
so rc =

SendNe tMessage (dwcl ien tS ess ions [ i ] , OPJREQUESTl ,0,0,0);

}

BP 0 649 121 A2

}

re turn (0) ;

}

int MessageHandlerl (unsigned long dwSessionld , PMSG pMsg)
{

int rc;

printf ("0P_REQUEST1 arrived #m (session %ld) \n" , dwSessionld)
pMsg->dwOpCode = OP_SERVICEl;
pMsg->dwSequence = 2 ;
pMsg->dwFlags = 0;
pMsg->dwBuffLen = 0;

rc = SendNetMessage(dwSessionId,OP_SERVICEl, 0,0,0) ;
printf ("Sending 0P_SERV1CE1 m» (%d)\n M , rc) ;
if (rc != SUCCESS)
{

while (rc == ERROR_CONNECT_IN_PROCESS )
{

DispatchNetMsg(O) ;

rc = SendNetMessage(dwSessionId,OP_SERVICEl ,0,0,0) ;
}

}

return (rc ) ;

int DefMsgHandler(unsigned long dwSessionld, PMSG pMsg)
return(O) ;

int ExceptionHandler(unsigned long dwSessionld, PMSG pMsg)
return(O) ;

int ClientHandlerl (unsigned long dwSessionld, PMSG pMsg)

EP 0 649 121 A2

{

printf 0P_SERVICE1 arrived (session %ld)\n" ,dwSessionId) ;

return(O) ;

)

The Grant conversation is originated by the Retail Accounting System 1504 of FIG. 15 in the store, most
likely as a result of a request to purchase originated by the customer at the PS. The structure of the message-
dependent section of the message passed back and forth between these two entities looks as follows:

ttdefine OP.GRANT 0x0170 // Command

#define OP_GRANTED 0x0171 // Response

^define OP_DENIED 0x0172 // Response

typedef struct _grantstruct

{

unsigned long dwOrderNum?

unsigned long dwStore;

unsigned long dwCIT;

unsigned long dwMedia;

char szFIlename [SIZE0F_ALBUM_NAME+1 ] •

char szGrantKey [SIZE0F_GRANT_KEY+1 ] ;

} GRANTSTRUCT, *PGRANT STRUCT ;

For example, a message sent from the Retail Accounting System 1504 to the Central Host 1506 could look
as follows:

GRANT ->

Op Code Sequence Flags Bufflen
00000170 00000001 FFFFFFFF 00000025

Buffer

76 54 32 45 54 01 02 12 64

EP 0 649 121 A2

And the return message (authorization) could look like:

5 GRANTED <-

Op Code Sequence Flags Bufflen Buffer

00000171 00000001 FFFFFF 00000025 54 34 98 76 54 01 90 23 20

Where the flags field has an unknown meaning at this point.
Had the request been denied the return message would look like:

DENIED <-

Op Code Sequence Flags Bufflen Buffer

00000172 00000001 FFFFFF 00000010 54 09 70 23 41

Where the flags field probably describes the reason for the denial.
25 The Manufacturing conversation is originated by the Retail Accounting System. The structure of the mes-

sage passed back and forth between the Retail Accounting 1504 and the Manufacturing Control Station 1508
looks as follows:

Note: These two structures are shared by both the Order and Manufacturing conversations.

ttdefine 0P_KAKE 0x0110 // Command

#define 0P_MADE 0x0111 // Response

35 #define 0P_CANN0T_MAKE 0x0112 // Response

typedef struct _itemstruct

40 I

unsigned long dwMedia,*
unsigned long dwCopies;
unsigned long dwTracks;
45 unsigned long dwCDBarcode;

unsigned long dwPricePerltem;

char szFileName [SIZEOF_ALBUM_NANE+l ] ;

char szAlbumTitle[SIZEOF_ALBUM_TTTLE+l ] ;

EP 0 649 121 A2

} ITEMSTRUCT, *PITEHSTRUCT f -

t.ypedef struct _orderstruct
f

unsigned long dwOrderNum;
unsigned long dwCustomer ID ;
unsigned long dwCIT;
unsigned long dwNumLinel terns ;
char szHowPaid[SIZEOF_HOW_PAIDU ]
ITEMSTRUCT I temBuf f [ 1 ] ;

} ORDERSTRUCT, *PORDERSTRUCT ;

For example, a message sent from the Retail Accounting Station to the Manufacturing Control Application
station 1508 could look as follows:

MAKE ->

Op Code Sequence Flags
00000110 00000001 FFFFFF

Bufflen Buffer
00000011 54 09 70 23 41

MADE <-
Op Code Sequence Flags
00000111 00000001 FFFFFF

Bufflen Buffer
00000011 54 09 70 23 41

40 The inability to manufacture would be signaled by a message with the following contents:

CANNOT_MAKE <-

Op Code Sequence Flags Bufflen Buffer

00000112 00000001 FFFFFF 00000011 54 09 70 23 41 ...

so Where the flags field indicates the reason why the requested item cannot be manufactured, and perhaps

also suggests a recommended action to remedy the problem. Problems such as device out of service, unable
to establish NetBios session, etc, are likely to be the reasons encoded in the flags field of the message.

The Stream conversations are more complex than the others since the conversation occurs in states such
as OPEN, READ, CLOSE.

EP 0 649 121 A2

ttdef ine

EXIST

0x0130

Command

ffuc i j. ne

ur.

_L A 1 D 1 3

fivfM ^1

U Jl V 1 Jl

Response

wderine

0P_

.OPEN

UXU1 J/.

Command

ftoe r ine

OP.

UXul J j

Response

ttdef ine

OP.

.SEEK

0x0134

Command

woe line

OP.

C IT Cl/ DtTClTT T

UXUlJ?

//Response

0r_

.KLAU

UXU 1 jo

Command

ttdef ine

0P_

_BUFF_READ

0x0137

Response

ttdef ine

0P_

.WRITE

0X0138

Command

ttdef ine

op_

.WRITTEN

0x0139

Response

ttdef ine

op_

.BLAST

0x0140

Command

ttdef ine

op.

.BLASTED

0x0141

Response

ttdef ine

op.

_XON

0x0142

Command

ttdefine

op.

.XOFF

0x0143

Command

ttdef ine

op.

.CLOSE

0x0144

Command

#define

0P_

.TRUNCATE

0x0145

Command

ttdefine

OP.

.TRUNCATED

0x0146

Response

ttdef ine

OP-

.STATUS

0x0147

Informational

(async)

typedef struct _streamstruct

{

30 unsigned long idCIT;

unsigned char Buffer[l];

} STREAMSTRUCT , *PSTREAMSTRUCT ;

For example, a conversation established between the PS and the Data Server or between the Device Con-
troller and the Data Server might look as follows:

EXIST ->

Op Code Sequence Flags Bufflen Buffer

00000130 00000001 FFFFFF 00000009 54 09 70 23 41 ...
EXISTS <-

EP 0 649 121 A2

Op Code Sequence Flags

00000131 00000001 FFFFFF

OPEN ->

Op Code Sequence Flags

00000132 00000001 FFFFFF

OPENED <-

Op Code Sequence Flags

00000133 00000001 FFFFFF

READ ->

Op Code Sequence Flags

00000136 00000001 FFFFFF

BUFF_READ <-

Op Code Sequence Flags

00000137 00000001 FFFFFF

SEEK ->

Op Code Sequence Flags

00000134 00000001 FFFFFF

SEEKJRESULT <-

Op Code Sequence Flags

00000135 00000001 FFFFFF

BLAST ->

Op Code Sequence Flags

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000140 00000001 FFFFFF

00000010 54 09 70 23 41

EP 0 649 121 A2

BLASTED <-

Op Code Sequence Flags

00000141 00000001 FFFFFF

XON ->

Op Code Sequence Flags

00000142 00000001 FFFFFF

XOFF ->

Op Code Sequence Flags

00000143 00000001 FFFFFF

TRUNCATE ->

Op Code Sequence Flags

00000145 00000001 FFFFFF

TRUNCATED <-

Op Code Sequence Flags

00000146 00000001 FFFFFF

WRITE ->

Op Code Sequence Flags

00000138 00000001 FFFFFF

WRITTEN <-

Op Code Sequence Flags

00000139 00000001 FFFFFF

CLOSE ->

Op Code Sequence Flags

00000144 00000001 FFFFFF

Bufflen Buffer
00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

EP0 649 121 A2

STATUS <->

Op Code Sequence Flags Bufflen Buffer

00000147 00000001 FFFFFF 00000010 54 09 70 23 41

The Order conversation can exist between the PS 1502 and the Retail Accounting Station 1504.

ttdefine OP_ORDER 0x0160 // Command

typedef struct _orderstruct

{

char szileName[9] ;

} ORDERSTRUCT , *PORDERSTRUCT;

For example, a message sent from the PS 1502 to the Retail Accounting Station 1504 during the process of
placing an order could look as follows:

ORDER ->

Op code Sequence Flags Bufflen Buffer
30 00000160 00000001 FFFFFF 00000009 FDS453FG0

The Device conversation takes place within the Manufacturing Control station 1508.

#def ine

0P_PREPARE

0x0180

Command

#define

0P_READY

0x0181

Response

ttdefine

0P_START

0x0182

Command

ttdefine

0P_STARTED

0x0183

Response

ttdefine

0P„RESTART

0x0184

Command

ttdefine

0P_RESTARTED

0x0185

// Response

ttdefine

0P_AB0RT

0x0186

Command

ttdefine

OP_ABORTED

0x0167

Response

ttdefine

0P_PAUSE

0x0168

Command

ttdefine

0P_PAUSED

0x0189

Response

ttdefine

0PJ)EVSTATUS

0x0190

Informational

ttdefine

OP_COMPLETED

0x0193

Informational

ttdefine

OP^SHUTDOWN

0x0194

Command

EP 0 649 121 A2

ttdefine 0P_D0WN 0x0195 // Response

5 typedef struct _devices truct

unsigned long dwStatusCode ; // return status codes

unsigned long dwDeviceType ; // device type

to unsigned long dwTrackCount ; // track (song) count

unsigned char szFileName ( 12 ] ; // audio/image file name

unsigned char szExtraBuf [ 1 ] ; // pointer to extra data

15 } DEVICESTRUCT, *PDEVICESTRUCT ;

20 For example, a typical conversation between the Manufacturing Control Application at station 1508 and a par-
ticular Device Controller may develop as follows:

OP_PREPARE ->

Op Code Sequence Flags

00000180 00000001 FFFFFF

0P_READY <-

Op Code Sequence Flags

00000181 00000001 FFFFFF

OP_START ->

Op Code Sequence Flags

00000182 00000001 FFFFFF

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

Bufflen Buffer

00000010 54 09 70 23 41

OP_STARTED <-
45 Op Code Sequence Flags

00000183 O0000001 FFFFFF

Bufflen Buffer
00000010 54 09 70 23 41

OP^RE START ->

Op Code Sequence Flags

Bufflen Buffer

00000184 00000001

OP_RE STARTED <-
Op Code Sequence

00000185 00000001

OP_ABORT ->
Op Code Sequence

00000186 00000001

OP_ABORTED <-
Op Code Sequence

00000187 00000001

OP_PAUSE ->
Op Code Sequence

00000188 00000001

OPJPAUSED <-
Op Code Sequence

00000189 00000001

OP_DEVSTATUS <-
Op Code Sequence

00000190 00000001
OP_COMPLETED <-

Op Code Sequence

00000193 00000001

OP_SHUTDOWN ->
Op Code Sequence

00000194 00000001

EP 0 649 121 A2